Brainstem bundles.
Detail from a figure in the study shows brainstem bundle regions of interest In three different cross-sections of a brain, as automatically determined by BSBT (top row) vs. a “ground-truth” method. Photo credit: Mark Olchanyi/MIT Picower Institute

The brainstem drives the body’s most essential functions — consciousness, breathing, heart rate and sleep — yet for decades, medical imaging has struggled to see it clearly.

Now, a team of researchers from MIT, Harvard and Massachusetts General Hospital has unveiled an AI-powered tool capable of mapping this “unexplored” region with unprecedented clarity.

In a study published in the Proceedings of the National Academy of Sciences, the team describes how their new software, the BrainStem Bundle Tool (BSBT), can resolve eight distinct nerve bundles that were previously invisible to standard MRI scanners.

“The brainstem is a region of the brain that is essentially not explored because it is tough to image,” says Mark Olchanyi, a doctoral candidate at MIT and the study’s lead author. “People don’t really understand its makeup from an imaging perspective.”

The motion problem

The challenge has always been one of movement and scale. The brainstem is small, complex, and constantly jiggled by the pulsing of blood and the rhythm of breathing, blurring the “white matter” cables that connect the brain to the body.

To overcome this, Olchanyi trained a “convolutional neural network” on high-resolution scans, teaching the AI to ignore the noise and identify the specific pathways of the nerve bundles.

“The brainstem is one of the body’s most important control centres,” says Emery N. Brown, a professor at MIT and co-senior author. “Mark’s algorithms… offer us new access to vital physiological functions such as control of the respiratory and cardiovascular systems.”

The potential for the tool was highlighted by the case of a 29-year-old man who spent seven months in a coma following a traumatic brain injury.

By applying the AI tool to scans taken throughout his recovery, doctors could see that his brainstem bundles had been displaced but not severed. As the man recovered, the software tracked the lesions shrinking by a factor of three and showed the nerve bundles physically moving back into their correct positions.

Diagnosing disease

Beyond trauma, the tool showed promise in diagnosing neurodegenerative conditions. When applied to scans of patients with Parkinson’s disease, multiple sclerosis (MS) and Alzheimer’s, the AI detected distinct patterns of damage.

For example, while Alzheimer’s patients showed decline in only one bundle, those with Parkinson’s showed degradation in three. In MS patients, the tool identified significant volume loss in specific pathways.

The researchers believe the tool could become a “key adjunct” for doctors, offering a non-invasive way to assess brainstem health and predict recovery.

Leave a Reply

Your email address will not be published. Required fields are marked *

You May Also Like

US team grabs aliens.gov as Reddit talks “Project Blue Beam” and disclosure

The US government has officially staked its claim to “aliens.gov,” igniting a…

AI bots can autonomously run massive propaganda campaigns as toxic teams

If you thought the bot networks that flooded social media during recent…

We must stop trusting big tech and start regulating AI to protect children online

From sexually explicit deepfakes to addictive social media algorithms, unregulated digital platforms…

The 2008 financial crash permanently downgraded American class identity

The Great Recession didn’t just devastate bank accounts — it fundamentally and…

Shrinking workforces threaten the global economy but AI can trigger huge growth

With birth rates plummeting and productivity flatlining, businesses are being forced to…

Neuroscience specialists discover seven types of psychological “hyperarousal”

Hyperarousal plays a crucial role in mental health, heavily influencing the severity…